support_new_device.html



<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
<head>
  <meta charset="utf-8">
  
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  
  <title>Design Doc: Supporting new Device/Library &mdash; PaddlePaddle  documentation</title>
  

    <link rel="stylesheet" href="../_static/css/theme.css" type="text/css" />
  

        <link rel="index" title="Index"
              href="../genindex.html"/>
        <link rel="search" title="Search" href="../search.html"/>
    <link rel="top" title="PaddlePaddle  documentation" href="../index.html"/> 
<script>
var _hmt = _hmt || [];
(function() {
  var hm = document.createElement("script");
  hm.src = "//hm.baidu.com/hm.js?b9a314ab40d04d805655aab1deee08ba";
  var s = document.getElementsByTagName("script")[0]; 
  s.parentNode.insertBefore(hm, s);
})();
</script>


  <script src="../_static/js/modernizr.min.js"></script>

</head>

<body class="wy-body-for-nav" role="document">

  <div class="wy-grid-for-nav">

    
    <nav data-toggle="wy-nav-shift" class="wy-nav-side">
      <div class="wy-side-scroll">
        <div class="wy-side-nav-search">
          

            <a href="../index_en.html" class="icon icon-home"> PaddlePaddle
          

          </a>

          
<div role="search">
  <form id="rtd-search-form" class="wy-form" action="../search.html" method="get">
    <input type="text" name="q" placeholder="Search docs" />
    <input type="hidden" name="check_keywords" value="yes" />
    <input type="hidden" name="area" value="default" />
  </form>
</div>

          
        </div>

        <div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
          
<nav class="doc-menu-vertical" role="navigation">

<ul>
<li class="toctree-l1"><a class="reference internal" href="../getstarted/index_en.html">GET STARTED</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../getstarted/quickstart_en.html">Quick Start</a></li>
<li class="toctree-l2"><a class="reference internal" href="../getstarted/concepts/use_concepts_en.html">Basic Concept</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../build_and_install/index_en.html">Install and Build</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/pip_install_en.html">Install using pip</a></li>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/docker_install_en.html">Run in Docker Containers</a></li>
<li class="toctree-l2"><a class="reference internal" href="../build_and_install/build_from_source_en.html">Build from Sources</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../howto/index_en.html">HOW TO</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../howto/cmd_parameter/index_en.html">Set Command-line Parameters</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/use_case_en.html">Use Case</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/arguments_en.html">Argument Outline</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cmd_parameter/detail_introduction_en.html">Detail Description</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/cluster/index_en.html">Distributed Training</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/preparations_en.html">Preparations</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/cmd_argument_en.html">Command-line arguments</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/cluster/multi_cluster/index_en.html">Use different clusters</a><ul>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/fabric_en.html">Fabric</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/openmpi_en.html">OpenMPI</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/k8s_en.html">Kubernetes</a></li>
<li class="toctree-l4"><a class="reference internal" href="../howto/cluster/multi_cluster/k8s_aws_en.html">Kubernetes on AWS</a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/rnn/index_en.html">RNN Models</a><ul>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/rnn_config_en.html">RNN Configuration</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/recurrent_group_en.html">Recurrent Group Tutorial</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/hierarchical_layer_en.html">Layers supporting hierarchical sequence as input</a></li>
<li class="toctree-l3"><a class="reference internal" href="../howto/rnn/hrnn_rnn_api_compare_en.html">API comparision between RNN and hierarchical RNN</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../howto/optimization/gpu_profiling_en.html">Tune GPU Performance</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../dev/index_en.html">Development</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../dev/contribute_to_paddle_en.html">Contribute Code</a></li>
<li class="toctree-l2"><a class="reference internal" href="../dev/write_docs_en.html">Contribute Documentation</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../faq/index_en.html">FAQ</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../faq/build_and_install/index_en.html">Install, Build and Unit test</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/model/index_en.html">Model Configuration</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/parameter/index_en.html">Parameter Setting</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/local/index_en.html">Local Training and Prediction</a></li>
<li class="toctree-l2"><a class="reference internal" href="../faq/cluster/index_en.html">Cluster Training and Prediction</a></li>
</ul>
</li>
</ul>

</nav>

        </div>
      </div>
    </nav>

    <section data-toggle="wy-nav-shift" class="wy-nav-content-wrap">

      
      <nav class="wy-nav-top" role="navigation" aria-label="top navigation">
        <i data-toggle="wy-nav-top" class="fa fa-bars"></i>
        <a href="../index_en.html">PaddlePaddle</a>
      </nav>


      <div class="wy-nav-content">
        <div class="rst-content">
          

<div role="navigation" aria-label="breadcrumbs navigation">
  <ul class="wy-breadcrumbs">
    <li><a href="../index_en.html">Docs</a> &raquo;</li>
      
    <li>Design Doc: Supporting new Device/Library</li>
      <li class="wy-breadcrumbs-aside">
        
          
            <a href="../_sources/design/support_new_device.md.txt" rel="nofollow"> View page source</a>
          
        
      </li>
  </ul>
  <hr/>
</div>
          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
           <div itemprop="articleBody">
            
  <div class="section" id="design-doc-supporting-new-device-library">
<span id="design-doc-supporting-new-device-library"></span><h1>Design Doc: Supporting new Device/Library<a class="headerlink" href="#design-doc-supporting-new-device-library" title="Permalink to this headline">¶</a></h1>
<div class="section" id="background">
<span id="background"></span><h2>Background<a class="headerlink" href="#background" title="Permalink to this headline">¶</a></h2>
<p>Deep learning has a high demand for computing resources. New high-performance devices and computing libraries are appearing very frequently. Deep learning frameworks have to integrate these high-performance devices and computing libraries in a flexible and efficient manner.</p>
<p>On one hand, hardware and computing libraries usually do not have a one-to-one correspondence. For example, Intel CPUs support Eigen and MKL computing libraries while Nvidia GPUs support Eigen and cuDNN computing libraries. We have to implement operator specific kernels for each computing library.</p>
<p>On the other hand, users usually do not want to care about the low-level hardware and computing libraries when writing a neural network configuration. In Fluid, <code class="docutils literal"><span class="pre">Layer</span></code> is exposed in <code class="docutils literal"><span class="pre">Python</span></code>, and <code class="docutils literal"><span class="pre">Operator</span></code> is exposed in <code class="docutils literal"><span class="pre">C++</span></code>. Both <code class="docutils literal"><span class="pre">Layer</span></code> and <code class="docutils literal"><span class="pre">Operator</span></code> are hardware independent.</p>
<p>So, how to support a new Device/Library in Fluid becomes a challenge.</p>
</div>
<div class="section" id="basic-integrate-a-new-device-library">
<span id="basic-integrate-a-new-device-library"></span><h2>Basic: Integrate A New Device/Library<a class="headerlink" href="#basic-integrate-a-new-device-library" title="Permalink to this headline">¶</a></h2>
<p>For a general overview of fluid, please refer to the <a class="reference external" href="https://github.com/PaddlePaddle/Paddle/blob/develop/doc/howto/read_source.md">overview doc</a>.</p>
<p>There are mainly three parts that we have to consider while integrating a new device/library:</p>
<ul class="simple">
<li>Place and DeviceContext: indicate the device id and manage hardware resources</li>
<li>Memory and Tensor: malloc/free data on certain device</li>
<li>Math Functor and OpKernel: implement computing unit on certain devices/libraries</li>
</ul>
<div class="section" id="place-and-devicecontext">
<span id="place-and-devicecontext"></span><h3>Place and DeviceContext<a class="headerlink" href="#place-and-devicecontext" title="Permalink to this headline">¶</a></h3>
<p>Please note that device and computing library are not one-to-one corresponding. A device can have a lot of computing libraries and a computing library can also support several devices.</p>
<div class="section" id="place">
<span id="place"></span><h4>Place<a class="headerlink" href="#place" title="Permalink to this headline">¶</a></h4>
<p>Fluid uses class <a class="reference external" href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/platform/place.h#L55">Place</a> to represent the device memory where data is located. If we add another device, we have to add the corresponding <code class="docutils literal"><span class="pre">DevicePlace</span></code>.</p>
<div class="highlight-default"><div class="highlight"><pre><span></span>        <span class="o">|</span>   <span class="n">CPUPlace</span>
<span class="n">Place</span> <span class="o">--|</span>   <span class="n">CUDAPlace</span>
        <span class="o">|</span>   <span class="n">FPGAPlace</span>
</pre></div>
</div>
<p>And <code class="docutils literal"><span class="pre">Place</span></code> is defined as follows:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">typedef</span> <span class="n">boost</span><span class="p">::</span><span class="n">variant</span><span class="o">&lt;</span><span class="n">CUDAPlace</span><span class="p">,</span> <span class="n">CPUPlace</span><span class="p">,</span> <span class="n">FPGAPlace</span><span class="o">&gt;</span> <span class="n">Place</span><span class="p">;</span>
</pre></div>
</div>
</div>
<div class="section" id="devicecontext">
<span id="devicecontext"></span><h4>DeviceContext<a class="headerlink" href="#devicecontext" title="Permalink to this headline">¶</a></h4>
<p>Fluid uses class <a class="reference external" href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/platform/device_context.h#L30">DeviceContext</a> to manage the resources in different libraries, such as CUDA stream in <code class="docutils literal"><span class="pre">CDUADeviceContext</span></code>. There are also inheritance relationships between different kinds of <code class="docutils literal"><span class="pre">DeviceContext</span></code>.</p>
<div class="highlight-default"><div class="highlight"><pre><span></span>                <span class="o">/-&gt;</span>  <span class="n">CPUDeviceContext</span>   
<span class="n">DeviceContext</span> <span class="o">----&gt;</span>  <span class="n">CUDADeviceContext</span>  
                \<span class="o">-&gt;</span>  <span class="n">FPGADeviceContext</span>
</pre></div>
</div>
<p>An example of Nvidia GPU is as follows:</p>
<ul class="simple">
<li>DeviceContext</li>
</ul>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">DeviceContext</span> <span class="p">{</span>
  <span class="n">virtual</span> <span class="n">Place</span> <span class="n">GetPlace</span><span class="p">()</span> <span class="n">const</span> <span class="o">=</span> <span class="mi">0</span><span class="p">;</span>
<span class="p">};</span>  
</pre></div>
</div>
<ul class="simple">
<li>CUDADeviceContext</li>
</ul>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">CUDADeviceContext</span> <span class="p">:</span> <span class="n">public</span> <span class="n">DeviceContext</span> <span class="p">{</span>
  <span class="n">Place</span> <span class="n">GetPlace</span><span class="p">()</span> <span class="n">const</span> <span class="n">override</span> <span class="p">{</span> <span class="k">return</span> <span class="n">place_</span><span class="p">;</span> <span class="p">}</span>
<span class="n">private</span><span class="p">:</span>
  <span class="n">CUDAPlace</span> <span class="n">place_</span><span class="p">;</span>
  <span class="n">cudaStream_t</span> <span class="n">stream_</span><span class="p">;</span> 
  <span class="n">cublasHandle_t</span> <span class="n">cublas_handle_</span><span class="p">;</span>
  <span class="n">std</span><span class="p">::</span><span class="n">unique_ptr</span><span class="o">&lt;</span><span class="n">Eigen</span><span class="p">::</span><span class="n">GpuDevice</span><span class="o">&gt;</span> <span class="n">eigen_device_</span><span class="p">;</span>  <span class="o">//</span> <span class="n">binds</span> <span class="k">with</span> <span class="n">stream_</span>
<span class="p">};</span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="memory-and-tensor">
<span id="memory-and-tensor"></span><h3>Memory and Tensor<a class="headerlink" href="#memory-and-tensor" title="Permalink to this headline">¶</a></h3>
<div class="section" id="memory-module">
<span id="memory-module"></span><h4>memory module<a class="headerlink" href="#memory-module" title="Permalink to this headline">¶</a></h4>
<p>Fluid provides the following <a class="reference external" href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/memory/memory.h#L36">memory interfaces</a>:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">template</span> <span class="o">&lt;</span><span class="n">typename</span> <span class="n">Place</span><span class="o">&gt;</span>
<span class="n">void</span><span class="o">*</span> <span class="n">Alloc</span><span class="p">(</span><span class="n">Place</span> <span class="n">place</span><span class="p">,</span> <span class="n">size_t</span> <span class="n">size</span><span class="p">);</span>

<span class="n">template</span> <span class="o">&lt;</span><span class="n">typename</span> <span class="n">Place</span><span class="o">&gt;</span>
<span class="n">void</span> <span class="n">Free</span><span class="p">(</span><span class="n">Place</span> <span class="n">place</span><span class="p">,</span> <span class="n">void</span><span class="o">*</span> <span class="n">ptr</span><span class="p">);</span>

<span class="n">template</span> <span class="o">&lt;</span><span class="n">typename</span> <span class="n">Place</span><span class="o">&gt;</span>
<span class="n">size_t</span> <span class="n">Used</span><span class="p">(</span><span class="n">Place</span> <span class="n">place</span><span class="p">);</span>
</pre></div>
</div>
<p>To implement these interfaces, we have to implement MemoryAllocator for different Devices.</p>
</div>
<div class="section" id="tensor">
<span id="tensor"></span><h4>Tensor<a class="headerlink" href="#tensor" title="Permalink to this headline">¶</a></h4>
<p><a class="reference external" href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/framework/tensor.h#L36">Tensor</a> holds data with some shape in a specific Place.</p>
<div class="highlight-cpp"><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">Tensor</span> <span class="p">{</span>
 <span class="k">public</span><span class="o">:</span>
  <span class="cm">/*! Return a pointer to mutable memory block. */</span>
  <span class="k">template</span> <span class="o">&lt;</span><span class="k">typename</span> <span class="n">T</span><span class="o">&gt;</span>
  <span class="kr">inline</span> <span class="n">T</span><span class="o">*</span> <span class="n">data</span><span class="p">();</span>

  <span class="cm">/**</span>
<span class="cm">   * @brief   Return a pointer to mutable memory block.</span>
<span class="cm">   * @note    If not exist, then allocation.</span>
<span class="cm">   */</span>
  <span class="k">template</span> <span class="o">&lt;</span><span class="k">typename</span> <span class="n">T</span><span class="o">&gt;</span>
  <span class="kr">inline</span> <span class="n">T</span><span class="o">*</span> <span class="n">mutable_data</span><span class="p">(</span><span class="n">platform</span><span class="o">::</span><span class="n">Place</span> <span class="n">place</span><span class="p">);</span>

  <span class="cm">/**</span>
<span class="cm">   * @brief     Return a pointer to mutable memory block.</span>
<span class="cm">   *</span>
<span class="cm">   * @param[in] dims    The dimensions of the memory block.</span>
<span class="cm">   * @param[in] place   The place of the memory block.</span>
<span class="cm">   *</span>
<span class="cm">   * @note      If not exist, then allocation.</span>
<span class="cm">   */</span>
  <span class="k">template</span> <span class="o">&lt;</span><span class="k">typename</span> <span class="n">T</span><span class="o">&gt;</span>
  <span class="kr">inline</span> <span class="n">T</span><span class="o">*</span> <span class="n">mutable_data</span><span class="p">(</span><span class="n">DDim</span> <span class="n">dims</span><span class="p">,</span> <span class="n">platform</span><span class="o">::</span><span class="n">Place</span> <span class="n">place</span><span class="p">);</span>

  <span class="cm">/*! Resize the dimensions of the memory block. */</span>
  <span class="kr">inline</span> <span class="n">Tensor</span><span class="o">&amp;</span> <span class="n">Resize</span><span class="p">(</span><span class="k">const</span> <span class="n">DDim</span><span class="o">&amp;</span> <span class="n">dims</span><span class="p">);</span>

  <span class="cm">/*! Return the dimensions of the memory block. */</span>
  <span class="kr">inline</span> <span class="k">const</span> <span class="n">DDim</span><span class="o">&amp;</span> <span class="n">dims</span><span class="p">()</span> <span class="k">const</span><span class="p">;</span>

 <span class="k">private</span><span class="o">:</span>
  <span class="cm">/*! holds the memory block if allocated. */</span>
  <span class="n">std</span><span class="o">::</span><span class="n">shared_ptr</span><span class="o">&lt;</span><span class="n">Placeholder</span><span class="o">&gt;</span> <span class="n">holder_</span><span class="p">;</span>

  <span class="cm">/*! points to dimensions of memory block. */</span>
  <span class="n">DDim</span> <span class="n">dim_</span><span class="p">;</span>
<span class="p">};</span>
</pre></div>
</div>
<p><code class="docutils literal"><span class="pre">Placeholder</span></code> is used to delay memory allocation; that is, we can first define a tensor, using <code class="docutils literal"><span class="pre">Resize</span></code> to configurate its shape, and then call <code class="docutils literal"><span class="pre">mutuable_data</span></code> to allocate the actual memory.</p>
<div class="highlight-cpp"><div class="highlight"><pre><span></span><span class="n">paddle</span><span class="o">::</span><span class="n">framework</span><span class="o">::</span><span class="n">Tensor</span> <span class="n">t</span><span class="p">;</span>
<span class="n">paddle</span><span class="o">::</span><span class="n">platform</span><span class="o">::</span><span class="n">CPUPlace</span> <span class="n">place</span><span class="p">;</span>
<span class="c1">// set size first</span>
<span class="n">t</span><span class="p">.</span><span class="n">Resize</span><span class="p">({</span><span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">});</span>
<span class="c1">// allocate memory on CPU later</span>
<span class="n">t</span><span class="p">.</span><span class="n">mutable_data</span><span class="p">(</span><span class="n">place</span><span class="p">);</span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="math-functor-and-opkernel">
<span id="math-functor-and-opkernel"></span><h3>Math Functor and OpKernel<a class="headerlink" href="#math-functor-and-opkernel" title="Permalink to this headline">¶</a></h3>
<p>Fluid implements computing units based on different DeviceContexts. Some computing units are shared between operators. This common part will be put in operators/math directory as basic Functors.</p>
<p>Let&#8217;s take <a class="reference external" href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/operators/math/maxouting.h#L27">MaxOutFunctor</a> as an example:</p>
<p>The interface is defined in the header file.</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">template</span> <span class="o">&lt;</span><span class="n">typename</span> <span class="n">DeviceContext</span><span class="p">,</span> <span class="n">typename</span> <span class="n">T</span><span class="o">&gt;</span>
<span class="k">class</span> <span class="nc">MaxOutFunctor</span> <span class="p">{</span>
 <span class="n">public</span><span class="p">:</span>
  <span class="n">void</span> <span class="n">operator</span><span class="p">()(</span><span class="n">const</span> <span class="n">DeviceContext</span><span class="o">&amp;</span> <span class="n">context</span><span class="p">,</span> <span class="n">const</span> <span class="n">framework</span><span class="p">::</span><span class="n">Tensor</span><span class="o">&amp;</span> <span class="nb">input</span><span class="p">,</span>
                  <span class="n">framework</span><span class="p">::</span><span class="n">Tensor</span><span class="o">*</span> <span class="n">output</span><span class="p">,</span> <span class="nb">int</span> <span class="n">groups</span><span class="p">);</span>
<span class="p">};</span>
</pre></div>
</div>
<p>CPU implementation is in .cc file</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">template</span> <span class="o">&lt;</span><span class="n">typename</span> <span class="n">T</span><span class="o">&gt;</span>
<span class="k">class</span> <span class="nc">MaxOutFunctor</span><span class="o">&lt;</span><span class="n">platform</span><span class="p">::</span><span class="n">CPUDeviceContext</span><span class="p">,</span> <span class="n">T</span><span class="o">&gt;</span> <span class="p">{</span>
  <span class="n">public</span><span class="p">:</span>
  <span class="n">void</span> <span class="n">operator</span><span class="p">()(</span><span class="n">const</span> <span class="n">platform</span><span class="p">::</span><span class="n">CPUDeviceContext</span><span class="o">&amp;</span> <span class="n">context</span><span class="p">,</span>
                  <span class="n">const</span> <span class="n">framework</span><span class="p">::</span><span class="n">Tensor</span><span class="o">&amp;</span> <span class="nb">input</span><span class="p">,</span> <span class="n">framework</span><span class="p">::</span><span class="n">Tensor</span><span class="o">*</span> <span class="n">output</span><span class="p">,</span>
                  <span class="nb">int</span> <span class="n">groups</span><span class="p">)</span> <span class="p">{</span>
                  <span class="o">...</span>
                  <span class="p">}</span>
<span class="p">};</span>
</pre></div>
</div>
<p>CUDA implementation is in .cu file</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">template</span> <span class="o">&lt;</span><span class="n">typename</span> <span class="n">T</span><span class="o">&gt;</span>
<span class="k">class</span> <span class="nc">MaxOutFunctor</span><span class="o">&lt;</span><span class="n">platform</span><span class="p">::</span><span class="n">CUDADeviceContext</span><span class="p">,</span> <span class="n">T</span><span class="o">&gt;</span> <span class="p">{</span>
 <span class="n">public</span><span class="p">:</span>
  <span class="n">void</span> <span class="n">operator</span><span class="p">()(</span><span class="n">const</span> <span class="n">platform</span><span class="p">::</span><span class="n">CUDADeviceContext</span><span class="o">&amp;</span> <span class="n">context</span><span class="p">,</span>
                  <span class="n">const</span> <span class="n">framework</span><span class="p">::</span><span class="n">Tensor</span><span class="o">&amp;</span> <span class="nb">input</span><span class="p">,</span> <span class="n">framework</span><span class="p">::</span><span class="n">Tensor</span><span class="o">*</span> <span class="n">output</span><span class="p">,</span>
                  <span class="nb">int</span> <span class="n">groups</span><span class="p">)</span> <span class="p">{</span>
                  <span class="o">...</span>
                  <span class="p">}</span>
<span class="p">};</span>                  
</pre></div>
</div>
<p>We first obtain the computing handle from a concrete DeviceContext and then compute on tensors.</p>
<p>The implementation of <code class="docutils literal"><span class="pre">OpKernel</span></code> is similar to math functors, the extra thing we need to do is to register the OpKernel in a global map.</p>
<p>Fluid provides different register interfaces in op_registry.h</p>
<p>Let&#8217;s take <a class="reference external" href="https://github.com/PaddlePaddle/Paddle/blob/develop/paddle/operators/crop_op.cc#L134">Crop</a> operator as an example:</p>
<p>In .cc file:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">REGISTER_OP_CPU_KERNEL</span><span class="p">(</span><span class="n">crop</span><span class="p">,</span> <span class="n">ops</span><span class="p">::</span><span class="n">CropKernel</span><span class="o">&lt;</span><span class="nb">float</span><span class="o">&gt;</span><span class="p">);</span>
<span class="n">REGISTER_OP_CPU_KERNEL</span><span class="p">(</span>
    <span class="n">crop_grad</span><span class="p">,</span> <span class="n">ops</span><span class="p">::</span><span class="n">CropGradKernel</span><span class="o">&lt;</span><span class="n">paddle</span><span class="p">::</span><span class="n">platform</span><span class="p">::</span><span class="n">CPUDeviceContext</span><span class="p">,</span> <span class="nb">float</span><span class="o">&gt;</span><span class="p">);</span>
</pre></div>
</div>
<p>In .cu file:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">REGISTER_OP_CUDA_KERNEL</span><span class="p">(</span><span class="n">crop</span><span class="p">,</span> <span class="n">ops</span><span class="p">::</span><span class="n">CropKernel</span><span class="o">&lt;</span><span class="nb">float</span><span class="o">&gt;</span><span class="p">);</span>
<span class="n">REGISTER_OP_CUDA_KERNEL</span><span class="p">(</span>
    <span class="n">crop_grad</span><span class="p">,</span> <span class="n">ops</span><span class="p">::</span><span class="n">CropGradKernel</span><span class="o">&lt;</span><span class="n">paddle</span><span class="p">::</span><span class="n">platform</span><span class="p">::</span><span class="n">CUDADeviceContext</span><span class="p">,</span> <span class="nb">float</span><span class="o">&gt;</span><span class="p">);</span>
</pre></div>
</div>
</div>
</div>
<div class="section" id="advanced-topics-how-to-switch-between-different-device-library">
<span id="advanced-topics-how-to-switch-between-different-device-library"></span><h2>Advanced topics: How to switch between different Device/Library<a class="headerlink" href="#advanced-topics-how-to-switch-between-different-device-library" title="Permalink to this headline">¶</a></h2>
<p>Generally, we will implement OpKernel for all Device/Library of an Operator. We can easily train a Convolutional Neural Network in GPU. However, some OpKernel is not suitable on a specific Device. For example, crf operator can only run on CPU, whereas most other operators can run on GPU. To achieve high performance in such circumstance, we have to switch between different Device/Library.</p>
<p>For more details, please refer to following docs:</p>
<ul class="simple">
<li>operator kernel type <a class="reference external" href="https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/operator_kernel_type.md">doc</a></li>
<li>switch kernel <a class="reference external" href="https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/switch_kernel.md">doc</a></li>
</ul>
</div>
</div>


           </div>
          </div>
          <footer>
  

  <hr/>

  <div role="contentinfo">
    <p>
        &copy; Copyright 2016, PaddlePaddle developers.

    </p>
  </div>
  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 

</footer>

        </div>
      </div>

    </section>

  </div>
  

    <script type="text/javascript">
        var DOCUMENTATION_OPTIONS = {
            URL_ROOT:'../',
            VERSION:'',
            COLLAPSE_INDEX:false,
            FILE_SUFFIX:'.html',
            HAS_SOURCE:  true
        };
    </script>
      <script type="text/javascript" src="../_static/jquery.js"></script>
      <script type="text/javascript" src="../_static/underscore.js"></script>
      <script type="text/javascript" src="../_static/doctools.js"></script>
      <script type="text/javascript" src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.0/MathJax.js?config=TeX-AMS-MML_HTMLorMML"></script>

  
    <script type="text/javascript" src="../_static/js/theme.js"></script>
  

  <script type="text/javascript">
      jQuery(function () {
          SphinxRtdTheme.StickyNav.enable();
      });
  </script>
   

</body>
</html>